Abstracts

Rationale: The repeated occurrence of seizures is the main feature of epilepsy. The mechanism underlying time-varying seizure risk is not known but is assumed to relate to time-varying cortical excitability. Many approaches that aimed to track time-varying seizure-risk have passively collected intracranial electroencephalographic recordings (iEEG), and applied algorithms for the identification of segments that associate with the pre-seizure state. Although passive observation of iEEG may allow some insights, active perturbation of the cortex and measuring the cortical response may provide much more direct information about cortical excitability. One way to do this would be to stimulate the cortex via intracranial electrodes and measure the cortical response using iEEG.

Methods: In this study, we analyse a cohort of eight epilepsy patients that were admitted to King’s College Hospital for presurgical intracranial EEG. During their stay, they underwent continuous single pulse electrical stimulation for 19 hours (mean, range 14-25) and stimuli were delivered every 5 minutes to a constant pair of electrodes. All patients experienced between 1 and 3 clinical seizures. In every patient the channel that showed the most prominent cortical response was chosen for analysis. After preprocessing the iEEG recordings (i.e., removal of the stimulation artifact, filtering) the data were epoched around the time of each stimulation. In each epoch we used the 5-100ms post-stimulation segment for analysis. We used the variance to quantify the stimulus response, and hence obtained a collection of variance values for all stimuli. Then, we normalized those values with the maximal value such that they vary from 0 to 1. For every seizure, we defined a 3-hour pre-ictal window, computed the average of the variance values and evaluated whether it was statistically significantly different from a distribution of 99 corresponding average values that were computed from 99 randomly selected 3-hour interictal epochs (one-sided randomization test; a = 0.05). The interictal epoch was defined as any 3-hour epoch that was at least 3-hour distant from the seizure onset and offset.

Results: In seven out of eight patients there was a significant change in the cortical response prior to the occurrence of seizures (p-values: [0.01;0.02]). In four patients there was a significant increase of the cortical response in the pre-ictal segment (Fig. 1a-d) and in three there was a decrease (Fig. 1e-g). In one patient there was not a significant change in the cortical response prior to seizure occurrence (Fig. 1h).

Conclusions: The findings suggest that the cortical excitability may change prior to seizure occurrence. Therefore, tracking the cortical response from electrical brain stimulation should be further explored as a biomarker for seizure susceptibility and aid in the development of seizure forecasting algorithms.

Funding: Please list any funding that was received in support of this abstract.: This study was supported by the NIHR, Biomedical Research Centre at South London and Maudsley NHS Foundation Trust; and by UK Medical Research Council ‘Developmental Pathway Funding’.